Slashdot Mirror
"Black Silicon" Advances Imaging, Solar Energy
waderoush writes "Forcing sulfur atoms into silicon using femtosecond laser pulses creates a material called 'black silicon' that is 100 to 500 times more sensitive to light than conventional silicon, in both the visible and infrared spectrums, according to SiOnyx, a venture-funded Massachusetts start-up that just emerged from stealth mode. Today's New York Times has a piece about the serendipitous discovery of black silicon inside the laboratory of Harvard physicist Eric Mazur. Meanwhile, a report in Xconomy explains how black silicon works and how SiOnyx and manufacturing partners hope to use it to build far more efficient photovoltaic cells and more sensitive detectors for medical imaging devices, surveillance satellites, and consumer digital cameras."
It's African-American Silicon, you insensitive clods!
Now I'm going to have to counteract this worrying news by expediting my research on black tinfoil.
This is another company using the mystique of "Trade secrets" to attract capital. If this is as good as they say, they wouldn't have any secrets and would spill the beans.
I think they have found some weaknesses that restrict the usefulness of this technology. Perhaps sensors made with this technology must be supercooled in order for them to function properly (i.e. perhaps this technology amplifies thermal noise by dozens of times).
Not all photons have the same energy (wavelength), and this is for precision imaging not power generation. Note it's more "sensitive" not more efficient.
"They were pure niggers." – Noam Chomsky
... again. I love solar power, and I realize that it progresses in small increments. But there have been so many stories of "break through" improvements that I don't really care until a profoundly more efficient product is made. Black silicon have twice the sensitivity to light that regular silicon does, which is great news for digital cameras and night vision scopes. I might be great news for solar power, but tell me about it once you have a working prototype with a noteworthy efficiency improvement.
We are all just people.
The efficiency of a solar cell is equal to the power absorbed by light divided by the power that is actually sent to the circuit the device is attached to. So if the sensitivity of the collector increases 500x, then there is likely going to be a major increase in the power supplied by the cell. This has nothing to do with the efficiency
Read carefully: they said 500x more sensitive than silicon, not 500x more sensitive than PV cells.
It's a bit like if they said that by reacting hydrogen with oxygen, they created a compound 700 times denser than oxygen. That doesn't mean it's 700 times denser than the densest material known.
Science writers who don't know what they are talking about annoy me,
There's an interesting irony to SiOnyx's business: a large chunk of the semiconductor industry's effort over the past 50 years has gone toward making silicon as pure as possible. But now SiOnyx and other companies are showing how useful--and perhaps profitable--it can be to craft silicon devices with impurities, defects, and unconventional structures.
A pure silicon crystal ingot and a doped silicon wafer are entirely different. You want a pure crystal to grow the ingot as large as possible. To make silicon useful you take the wafer sliced form the ingot, ant it has to be doped (ie add impurities) amongst many other steps.
LetterRip
How many of them thought to utilize angry sea bass in the process? Hmm? Makes you wonder doesn't it?
I might be great news for solar power, but tell me about it once you have a working prototype with a noteworthy efficiency improvement.
From what I've read this story is more about image sensors, but for solar cell applications: I don't understand the fuss about all these 'breakthrough efficiency record' stories. For all but a few applications (think satellites, pocket calculators etc.) efficiency doesn't matter. There is no shortage of sunlight, and therefore no need to turn a maximum of it into electricity. What matters is price per generated electric power ($/Watt), and how long the solar cells will last.
If I'm not mistaken, the solar cell market is hitting the 1 $/Watt mark around now, and growing at what, 10% ? 20% ? 50% per year? Wake me up when solar cells become cheaper than roof tiles, or provide a return on investment in <5 years (for average households), and will last decades after that. Then you have a breakthrough.
But there have been so many stories of "break through" improvements that I don't really care until a profoundly more efficient product is made.
Some years back, I read an article in an old magazine (I think it was a 1960's Popular Science) about a new method of blowing glass resulting in "near unbreakable" bottles. It went on excitedly for page, after page, talking about the new era of safety that this kind of glass could behest - glass that doesn't easily break - you could drop your soda or medicine bottle and it wouldn't shatter!
Intrigued, I spent an entire afternoon at the local University library trying to figure out exactly what happened to this miraculous technology! I even did some searching (AltaVista) on the then new-fangled Internet. The truth rather surprised me...
This "breakthrough" technology that had gone invisible was part of my everyday life, including the bottle of Diet Coke I was then slurping from! It had become so common that virtually nobody produced the old-fashioned fragile bottles and glass anymore!
That's why it works to have coffee tables with glass counter tops. That's why restaurants can get away with the sterile, easily cleaned, hard-to-scratch glass overlays on their tables. Next time you are at a corner market and see the glass countertop with the items for sale inside, think about that article in the ancient Popular Science article.
Once breakthroughs actually become available, they don't seem like breakthroughs - they quickly just become part of the landscape, and people don't notice them, anymore. This is why the "Intelligent Design" idiots can get out of their incredibly complex, affordable, high-tech SUVs and then announce that Science has it all wrong. Once it's routine, it no longer seems like such a big deal.
Proof? Affordable, thin-film photovoltaics is still largely considered a "breakthrough" technology. But there's a company doing it now, today, affordably. Alas, while they are growing as fast as they are able, all their production capacity is already sold to germany. I'd suggest you read up on it.
High tech is introduced slowly. At first, the high engineering cost can only be paid in niche markets where the return on investment is fat. But as the original engineering cost gets paid back, and as the technology itself is matured and tested, the cost of implementation drops rapidly, so that it applies to more and more and more niches. By the time it's available for common Joes like you and me, it doesn't seem like such a big deal, and we are left wondering "where are the breakthroughs?" from our satellite/GPS navigated, MP3 playing, fuel-injected, ABS-brakes protecting, vulcanized rubber-tired, air-conditioned, hybrid gas/electric, high-tech wonder machine.
Where are the breakthroughs? Look at the beer bottle in your trashcan.
I have no problem with your religion until you decide it's reason to deprive others of the truth.
We'll effectively be stealing energy that DIDN'T strike earth! Sucks to be anyone else harvesting energy from Sol...
Just to be a bit more explicit, sensitivity probably refers to one of two things.
The first would be an increase in quantum efficiency; that would be an increase in the ratio of photons detected to those impacting. In a photovoltaic cell this would lead to improved efficiency. Current scientific detectors, that I've looked into anyway for a research project I'm involved in, max out at maybe 70%, with most reasonably priced ones being 25%-35%. (The 70% ones tend to be things like photomultiplier tubes which require power input to achieve a high reverse voltage, so they're certainly not useful for PV cells.)
The second aspect would be to decrease the noise or dark count so that its capable of detecting dimmer and dimmer light sources, and in order to get the > 100% improvements this is definitely a large aspect of what the new method has done. Unfortunately I know more about the applications and figures of merit than the semiconductor stuff, so I can't say much about this other than I hope this opens up some new application possibilities.
The problem with most uncooled imagers isn't insufficient sensitivity any more. It's thermal noise. Unless this improves the S/N ratio, it won't help for uncooled imagers. That's why digital cameras which increase sensitivity in darkness show more and more noise as less light is received.
Cooled imagers, though, as in astronomy and fancier night vision equipment, might benefit. Cooling is done to reduce the random photons from heat within the imager. So cooled imagers do run into the sensitivity limitations of silicon, and might benefit.
But that's an exotic application. Cooled imagers are found mostly in military, space, and astronomy. Some require liquid nitrogen. It's not a mainstream technology.
If this is as good as they say, they wouldn't have any secrets and would spill the beans.
The fundamental research was done a long time ago(with picture of prototypes); I've read articles about it in Electronics and Wireless World several times over the years, so it's hardly a secret. Any potentially patentable critical element is going to be kept under wraps, obviously.
I think they have found some weaknesses that restrict the usefulness of this technology.
Or they spent 3 years on R&D fixing those weaknesses, like the article says.
Further information of note from the NYT article:
So we're told:
1- There's a decade of peer-reviewed research behind the technology.
2- They have funding and partners already.
3- They're shipping parts now, not at some unknown time in the future.
Either this is real, or Dr Mazur et al are engaging in an exceptionally elaborate, very public and career-ending series of lies (and it's not as though SiOnyx will be a paying proposition if the tech doesn't work). The part of the operation that does look suspect is their web site (Flash warning), but that doesn't prove anything about the physics involved.
Blank until
I'm sorry, I fail to see where the car fits in.
Semi-automatic amateur armchair Australian philosopher; conjecture ready at any moment...
You are right, that is the idea. From Behe's book "Darwin's Black Box" (a pretty stupid book) the 'problem' is that systems can be "irreducibly complex". That is, like the mousetrap - remove or change any part and it stops working.
The problem again (and since Behe's is a biochemist he is either stupid or lying if he doesn't understand this) is that nature builds her mousetraps in a very different way.
All previous 'versions' of any particular mousetrap (or other design) HAD to work. Small changes to them, including replacing parts or modifying parts were made, and those mousetraps that failed to catch any mice were rejected (died off).
This is only possible with systems whose parts can be replaced with slightly similar ones, and still sort of work. Evolutionary systems have evolved to be evolvable.
So, it's not the self-assembly, but the mutability of natural systems that is under dispute. Most biologists understand that natural systems can change quite radically - species evolution - while a few ID-ers just don't get it. Their loss; natural systems are truly astonishing.
They already spilled the beans - femtosecond laser pulses against silicon wafer in sulfur hexafluoride gas.
The problems probably are:
1. femtosecond laser pulses aren't exactly easy to make
2. the power density of the beam (if they increase the spot size, the power density goes down, meaning it's more costly and difficult to expose larger portions of the wafer at once, hence increasing time and cost)
3. sulfur hexafluoride - ummm hexafluoride anything is probably not the safest thing to deal with, hence - increasing cost
4. effects of oxide formation post-processing probably increases problems
5. thermal noise ... probably not much of an issue, plus I don't think they're talking about far IR photons, just IR that would normally be picked up in a GaAs detector
6. there is no mention of what wavelength of laser light they're using, so if it's something in the UV range, they'd need more expensive optics, increasing costs yet again.
I just want these SiOnyx people to do this with Uranium Hexafluoride. I want them to do it NOAW!
That should have a beautifully large cross-section, gobbling up lots of photons, and would give the nuclear industry something to do besides fission. This is just a guess, maybe there's some fucked up reason UF6 just wouldn't work for this purpose. I just like the idea of increasing the absorption band of photovoltaics.
meh ... back to reading the Modern Physics textbook.
The company name is a good choice, it sounds like Psionics, which implies of Psi, the Greek letter used to represent wavefunctions, on top of the onyx for black.
I'm one step closer to psionic powers!
:(
Wait, nevermind, they said SiOnyx.
If you can read this, I forgot to post anonymously.
If you read the journal articles http://dx.doi.org/10.1016/j.mseb.2006.10.002 you'll find that this process esentially creates a large number of impurity states at the center of the band gap, creating an impurity band. What this means is that light is absorbed very very fast, but then its also turned to to heat very very fast. In other words you can excite electrons but that electron will decay back down before it creates any current. This could still work for a photodetector because you can apply a voltage to sweep out the excited carriers before they recombine/decay but not for a solar cell since you want to generate power.
The sensitivity they are referring to is the amount of electrons released by the incident light - Amps of current per Watt of sunlight. Sunlight has a broad spectrum, and this technique allows more of the infrared portion of the spectrum (which is a lot) to cause electrons to flow.
However, and this is important, they achieved this by lowering the bandgap energy of the silicon. Why is that important? Remember that power, when it comes to electronics, is current times voltage. The voltage of a solar cell (open circuit voltage) is more or less the bandgap energy (divided by one electron charge). So, yeah, they get more electrons to flow for the same amount of incident sunlight, but the cell's voltage has also been lowered. Do you end up with more or less power as a result? Does the greater current overcome the lowered voltage? Since they haven't actually published data on a solar cell made from this technique, there isn't really a way to tell for certain.
My guess is that they won't be able to get vast power gains - possibly lower ones. The reason for this is that, right now, one photon with energy greater than the bandgap energy has a chance to create one electron-hole pair. If the photon has more energy than the bandgap energy, it doesn't make a correspondingly more energetic electron-hole pair. Even if the photon had twice the bandgap energy, it can't make two electron-hole pairs. So, a blue photon creates as much useful electrical energy as a red photon, despite the fact that the blue photon has more energy in it. One can play around with the bandgap energy of the PV cell to make better use of the high energy photons, but at the cost of excluding lower energy photons like infrared and red. More info here. This is why the solar cells with greatest efficiency are actually multi-junction cells - several solar cells with different bandgap energies stacked on top of each other, each tuned to a different portion of the solar spectrum.
The article mentions how these guys should be able to use their black silicon to create multiple electron-hole pairs from a single photon. In order to do that, however, they have to provide a bias voltage. In that case, the solar cell is sucking power, not producing it. That's fine if what you want is a very sensitive photo sensor - it's basically a solid-state photomultiplier tube. It's not a way to generate electrical power.
The sensors in cameras are already many many times more sensitive to light then the rods in your eyes.
Sensors exist that can detect single photons (if properly cooled). However the sensors are not as flexible as the human eye. They tend to have a linear response to light intensity rather then the eyes Log response (rods and cones don't actually respond to the intensity of light but the signal generated by a change seems to be log) although some sensors exist that produce log outputs.
Sensors are hitting thier sensitivity limits (think low light photography) but that is in terms of sensor noise. The human eye has far more noise to but it has a massive chunk of grey matter behind it that really helps filter the noise out.
Another difference is that the cones in you eye do not respond to the intensity of light they respond to the change in the intensity of light. Sensors respond directly to the intensity of light.
So improving the sensitivity of a sensor is unlikely to have any much impact on normal photography, perhaps it may reduce the dark noise.